Intercalated Cell BK-/4 Channels Modulate Sodium and Potassium

BASIC RESEARCH www.jasn.org

Intercalated Cell BK-␣/␤4 Channels Modulate Sodium and Potassium Handling During Potassium Adaptation

J. David Holtzclaw, P. Richard Grimm, and Steven C. Sansom

Department of Cellular and Integrative Physiology, University of Nebraska Medical Center, Omaha, Nebraska

ABSTRACT The large-conductance, calcium-activated potassium (BK) channels help eliminate potassium in mammals consuming potassium-rich diets. In the distal nephron, principal cells contain BK-␣/␤1 channels and intercalated cells contain BK-␣/␤4 channels. We studied whether BK-␤4–deficient mice Ϫ Ϫ (Kcnmb4 / ) have altered renal sodium and potassium clearances compared with wild-type mice when fed a regular or potassium-rich diet for ten days. We did not detect differences in urinary flow or fractional excretions of potassium (FEK) or sodium (FENa) between Kcnmb4-deficient and wild-type mice fed a regular diet. However, a potassium-rich diet led to Ͼ4-fold increases in urinary flows for both groups of mice, although Kcnmb4-deficient mice exhibited less urinary flow, higher plasma potassium concentration, more fluid retention, and significantly lower FEK and FENa than wild-type mice despite similar plasma aldosterone levels. Immunohistochemical analysis revealed increased basolateral Na-K- ATPase in principal cells of all potassium-adapted mice, but expression of Na-K-ATPase in intercalated cells was Ͼ10-fold lower. The size of intercalated cells reduced and luminal volume increased among potassium-adapted wild-type but not Kcnmb4-deficient mice. Paradoxically, this led to increased urinary fluid velocity in potassium-adapted Kcnmb4-deficient mice compared with wild-type mice. Taken to- gether, these data suggest that BK-␣/␤4 channels in intercalated cells reduce cell size, increasing luminal volume to accommodate higher distal flow rates during potassium adaptation. These changes streamline flow across the principal cells, producing gradients more favorable for potassium secretion and less favorable for sodium reabsorption.

J Am Soc Nephrol 21: 634–645, 2010. doi: 10.1681/ASN.2009080817

A high-K diet is a natural diuretic,1 causing decreased ate Na and water reabsorption and K secretion, Na and Cl reabsorption in the thick ascending limb and the ICs mediate acid/base transport. Under (TAL) because of medullary recycling and high inter- normal conditions, K secretion by the PCs is me- stitial K levels.2 The decreased Na transport in the diated primarily by the ROMK channel.7 How- medullary TAL disrupts the concentrating mecha- ever, flow-induced K secretion in the distal nism, thereby increasing flow to the distal nephron.2 nephron is mediated by BK.4,8,9 The high deliveries of Na to the connecting tubules BK are a complex of pore-forming ␣ and acces- (CNT) and cortical collecting ducts (CCD) is ex- sory ␤ subunits (BK-␣/␤). The PCs of the CNT changed for K, and the increased flow stimulates K secretion to maximize the amount of K secreted to Na absorbed. The renal outer medullary kidney K chan- Received August 11, 2009. Accepted December 11, 2009. nel (ROMK) and the large conductance, calcium-ac- Published online ahead of print. Publication date available at www.jasn.org. tivated K channels (BK) in the CNT and CCD serve to eliminate K during K adaptation.3–6 Correspondence: Dr. Steven C. Sansom, Department of Cellular and Integrative Physiology, 985850 Nebraska Medical Center, In the distal nephron, the CNT and CCD con- Omaha, NE 68198-5850. Phone: 402-559-2919; Fax: 402-559- sist of two epithelial cell types: principal cells 4438; E-mail: [email protected] (PCs) and intercalated cells (ICs). The PCs medi- Copyright ᮊ 2010 by the American Society of Nephrology

634 ISSN : 1046-6673/2104-634 J Am Soc Nephrol 21: 634–645, 2010 www.jasn.org BASIC RESEARCH contain BK-␣/␤1 and are well equipped with an abundance of significantly lower compared with KA WT (P Ͻ 0.001) and Ϫ Ϫ basolateral Na-K-ATPase to secrete K in K-adapted (KA) con- control Kcnmb4 / (0.39 Ϯ 0.02%, P Ͻ 0.02). The decrease in ditions. However, the preponderance of BK-␣ reside in ICs10,11 Na excretion indicated a large increase in Na reabsorption in along with the ancillary subunit, BK-␤4 (gene: Kcnmb4).12 A the distal nephron. Ϫ Ϫ study has indicated that BK of ICs are regulated by mitogen- As expected, the KA WT and KA Kcnmb4 / exhibited a activated protein kinase to prevent K reabsorption during de- significantly greater fractional excretion of K (FEK) compared Ϫ/Ϫ mand for maximal K secretion.13 It has also been proposed that with controls (Figure 1B). However, FEK for KA Kcnmb4 Ϯ BK-␣/␤4 in ICs have a role in flow-mediated K secretion. (93.5 10.8) was significantly attenuated compared with KA Ϯ Ͻ If high flow induces BK-mediated K secretion, then BK-␣/␤4 WT (162.6 10.8%, P 0.001). There was no statistical dif- Ϫ/Ϫ of ICs must have a role. It has been shown that the shear stress ference in FEK between control WT and control Kcnmb4 Ϯ Ϯ ϭ produced by high flow causes a transient increase in intracellular (23.3 1.5% versus 24.0 0.4%, P 0.967). There was no Ca to levels that may activate BK. Indeed, ICs, which protrude into statistical difference in GFR between groups (data not shown). the lumens of the CNT and CCD, are particularly subjected to The plasma Na and K concentrations are shown in Figure 1, C and D, respectively. The plasma Na levels were not signifi- shear stress forces that may elevate intracellular Ca. However, a cantly different among the four treatment groups, with values transient Ca activation of BK would not produce the sustained K varying from 137 Ϯ 1 mM for control WT to 139 Ϯ 2 mM for transport required for long-term K adaptation. Ϫ Ϫ KA Kcnmb4 / (Figure 1C). The plasma K concentrations of That BK-␣/␤4 of ICs may not produce sustained K secre- Ϫ Ϫ control WT and control Kcnmb4 / were not different (Fig- tion is also indicated by the paucity of Na-K-ATPase. K adap- ure 1D). The plasma K concentrations for KA WT was slightly 14 15–17 tation or mineralocorticoid treatment increases the but insignificantly (P ϭ 0.085) increased compared with con- quantity of basolateral Na-K-ATPase of mammalian collecting trol WT (4.33 Ϯ 0.11 versus 4.05 Ϯ 0.05 mM). The plasma K ducts to maintain a favorable electrochemical driving force for concentration of KA Kcnmb4Ϫ/Ϫ (4.75 Ϯ 0.12 mM) was sig- K secretion. However, because ICs have considerably less Na- nificantly greater than Kcnmb4Ϫ/Ϫ control (4.03 Ϯ 0.02 mM; K-ATPase than PCs,18–21 ICs may not have an adequate K P Ͻ 0.001) and KA WT (P Ͻ 0.005). This increase in plasma K ␣ ␤ 22 source to sustain K secretion via BK- / 4. Still, Na-K-AT- concentration reflected the attenuated K secretory response of Pase has not been quantified in ICs in KA conditions. If the KA Kcnmb4Ϫ/Ϫ. BK-␣/␤4 were directly involved in the increased K transport Urinary output and water intake (by drinking) are shown in associated with K adaptation, then it would be expected that Figure 2. As shown in Figure 2A, the urinary output of control the Na-K-ATPase in ICs would increase summarily to PCs.23 Kcnmb4Ϫ/Ϫ (1.01 Ϯ 0.02 ml/d) was not significantly different Ϫ Ϫ To this end, we determined whether Kcnmb4 / have al- from control WT (1.03 Ϯ 0.03 ml/d). The urinary output in- tered renal K and Na excretions compared with wild type (WT) creased in KA WT by nearly 5-fold to 4.94 Ϯ 0.15 ml/d. The under control and KA conditions. Evidence from KA Kc- urinary output increased in KA Kcnmb4Ϫ/Ϫ to 4.32 Ϯ 0.14 nmb4Ϫ/Ϫ indicates that the role of BK-␣/␤4 in ICs is to reduce ml/d; however, this value was significantly attenuated com- cell size, thereby increasing tubular fluid volumes to accom- pared with KA WT. modate the higher distal flow rates of KA mice. By reducing the As shown in Figure 2B, the water consumption of control protrusion of ICs into the lumen and increasing tubular vol- KA (1.93 Ϯ 0.07 ml/d) was not different from control WT ume, flow will be more streamlined across the PCs and a more (1.87 Ϯ 0.19 ml/d). The water consumption of KA WT and KA Ϫ/Ϫ Ϫ/Ϫ a favorable chemical gradient for K secretion and less favorable Kcnmb4 increased substantially; however, KA Kcnmb4 Ϯ gradient for Na reabsorption will be produced. consumed significantly more water (7.70 0.17 ml/d) compared with KA WT (6.42 Ϯ 0.14 ml/d). Therefore, consistent with Na retention, KA Kcnmb4Ϫ/Ϫ consumed more water and had less urinary output compared with KA WT. RESULTS Figure 3 shows volume status as related to weight changes (Figure 3A) and hematocrits (Figure 3B) for the four treatment Na, K, and Volume Balance groups. The KA WT gained slightly more weight (0.42 Ϯ Experiments were performed to determine the difference in Na 0.03 g) than control WT (0.22 Ϯ 0.03 g). However, KA Kc- and K handling and volume balance in four groups of mice: nmb4Ϫ/Ϫ gained 5.2 Ϯ 0.2 g, which was significantly greater control diet wild type (control WT), control diet Kcnmb4Ϫ/Ϫ than the weight gain of KA WT (P Ͻ 0.001) and of control (control Kcnmb4Ϫ/Ϫ), K-adapted wild type (KA WT), and K- Kcnmb4Ϫ/Ϫ (0.21 Ϯ 0.02 g; P Ͻ 0.001). The hematocrits of KA adapted Kcnmb4Ϫ/Ϫ (KA Kcnmb4Ϫ/Ϫ). A sample number (N) and control WT were not significantly different (45.7 Ϯ 0.4 between 7 and 11 represented all groups. versus 45.5 Ϯ 0.4%). However, the hematocrits of KA Kc- Ϫ/Ϫ As shown in Figure 1A, the fractional excretion of Na (FENa) nmb4 were significantly lower compared with control Kc- was significantly (P Ͻ 0.001) increased for KA WT (0.60 Ϯ nmb4Ϫ/Ϫ (35.4 Ϯ 0.5% versus 45.9 Ϯ 0.4%; P Ͻ 0.001). The 0.03%) compared with control WT (0.40 Ϯ 0.02%). However, increased weight and lower hematocrit reveal extreme fluid Ϫ/Ϫ Ϯ Ϫ/Ϫ the FENa of KA Kcnmb4 was only 0.29 0.03%, a value retention and volume expansion for KA Kcnmb4 .

J Am Soc Nephrol 21: 634–645, 2010 Role for IC in Na and K Handling 635 BASIC RESEARCH www.jasn.org

A B

C D

Ϫ/Ϫ Figure 1. Kcnmb4 mice have altered K and Na handling under K adaptation. Bar plots showing (A) FeNa and (B) FEK and plasma Ϫ Ϫ concentrations for (C) Na and (D) K for WT and Kcnmb4 / control and KA conditions. Data represent mean Ϯ SEM. * ϭ significant difference (P Ͻ 0.05) between diets. † ϭ significant difference between phenotypes on same diet.

Aldosterone Compared with control WT (96.9 Ϯ 1.8) and control Kc- Figure 4 shows the plasma aldosterone levels for control and nmb4Ϫ/Ϫ (92.0 Ϯ4.9), KA WT and KA Kcnmb4Ϫ/Ϫ exhibited KA WT and Kcnmb4Ϫ/Ϫ. The aldosterone levels were not sig- significantly greater intensity values of 157.5 Ϯ 5.5 and 146.7 Ϯ nificantly different in control Kcnmb4Ϫ/Ϫ (116.0 Ϯ 1.1 pg/ml) 9.3, respectively. There was no significant difference in PC Na-K- compared with control WT (106.6 Ϯ 2.7 pg/ml). As expected, ATPase expression between control WT and control Kcnmb4Ϫ/Ϫ K adaptation resulted in significantly increased plasma aldo- or between KA WT and KA Kcnmb4Ϫ/Ϫ. sterone concentrations for WT and Kcnmb4Ϫ/Ϫ. However, the As shown in Figures 5A through 5D, the Na-K-ATPase ex- plasma aldosterone concentration for KA Kcnmb4Ϫ/Ϫ was pression on ICs was barely detectable, with relative staining 149.4 Ϯ 6.9 pg/ml, a value that was not significantly different values Ͻ10% of PC staining values. There was no significant from the value of 143.8 Ϯ 9.9 pg/ml for KA WT. difference in IC Na-K-ATPase staining when control was compared with KA conditions. In addition, there was no difference Differential Effects of K-Adaptation on Na-K-ATPase when IC Na-K-ATPase expression of KA WT was compared Ϫ Ϫ of ICs and PCs with that of KA Kcnmb4 / (15.97 Ϯ 3.65 versus 13.80 Ϯ Immunohistochemical co-staining for V-ATPase (red), a marker 1.82). These results demonstrate that PCs, but not ICs, are ␣ of ICs, and the 1 subunit of the Na-K-ATPase (green) is shown in adapting to a high-K diet by elevating Na-K-ATPase, which Figures 5A through 5D and is quantified for with relative expres- increases the driving force for K secretion. Moreover, the de- sion units (AU) for PCs and ICs in each of the four groups (n ϭ 3 crease in K secretion in KA Kcnmb4Ϫ/Ϫ cannot be attributed to in each group). For the CCD of control WT and Kcnmb4Ϫ/Ϫ, the a deficiency of Na-K-ATPase in PCs or ICs. Na-K-ATPase stained faintly on the basolateral membrane (BLM) of ICs and intensely on the BLM of PCs (absence of V- Immunohistochemical Analysis of Tubular Cross- ATPase; non-IC). The summary bar graph of Figure 5E compares Sectional Area and IC Volume the PC Na-K-ATPase intensities (Figure 5E) of the four groups. Despite the decrease in K excretion in Kcnmb4Ϫ/Ϫ, the mini-

636 Journal of the American Society of Nephrology J Am Soc Nephrol 21: 634–645, 2010 www.jasn.org BASIC RESEARCH

Ϫ Ϫ Figure 3. Kcnmb4 / mice undergo volume expansion during

Ϫ/Ϫ KA. Bar plots illustrating fluid balance as related to (A) weight Figure 2. Kcnmb4 mice retain water during KA. (A) Urine Ϫ Ϫ change and (B) hematocrit in WT and Kcnmb4 / in control and output and (B) water intake values for each treatment group. * ϭ KA conditions. * ϭ significant difference (P Ͻ 0.05) between diets. significant difference (P Ͻ 0.05) between diets. † ϭ significant † ϭ significant difference between phenotypes on same diet. difference between phenotypes on same diet. mal levels of Na-K-ATPase in ICs of KA mice suggest a nontran- sepithelial transport role for BK-␣/␤4. Because ICs normally protrude into the lumen of the CNT and CCD, we questioned whether the BK-␣/␤4 would have a volume regulatory role during the high flow–high shear stress of KA conditions to help accommodate the larger fluid volume. If so, the lumens of KA Kc- nmb4Ϫ/Ϫ would be narrowed compared with KA WT. The representative colorimetric immunohistochemical photomicro- graphs of CCD are shown in Figures 6A through 6D, and a summary of tubular cross-sectional areas are shown for CCD (Figure 6E) and CNT (Figure 6F). There were significant increases in CCD and CNT cross-sectional areas of KA WT when compared Ϫ/Ϫ Figure 4. KA increases plasma aldosterone. Plasma aldosterone with control WT. KA Kcnmb4 also exhibited significantly el- Ϫ Ϫ levels for WT and Kcnmb4 / on control and KA diets. * ϭ evated cross-sectional areas of the CNT and CCD compared with significant difference (P Ͻ 0.05) between diets. † ϭ significant Ϫ/Ϫ Ͻ control Kcnmb4 (P 0.005). However, these values were sig- difference between phenotypes on same diet. nificantly less than KA WT. There were no statistical differences in cross-sectional areas of CNT or CCD when comparing control hibit cell volume decrease. We therefore placed WT and Kc- WT with control Kcnmb4Ϫ/Ϫ. These data show that CCD and nmb4Ϫ/Ϫ on a control or high-K diet and determined whether CNT tubules have increased luminal volumes in response to K- ICs exhibited cell volume reduction in the presence or absence adaptation-induced fluid flow and this increased luminal volume of the BK-␤4 in vivo. IC volume was estimated using the well Ϫ Ϫ is blunted in Kcnmb4 / . established Cavalieri method.24 Summary bar plots of IC vol- If flow-induced shear stress activates BK-␣/␤4 in the ab- umes are shown for CCD (Figure 7A) and CNT (Figure 7B). sence of matching K delivery by Na-K-ATPase, ICs would ex- Under the high flow, high shear stress conditions of a high-K

J Am Soc Nephrol 21: 634–645, 2010 Role for IC in Na and K Handling 637 BASIC RESEARCH www.jasn.org

Ϫ Ϫ Figure 5. KA increases Na-K-ATPase expression in PC. Representative sections from (A) WT control, (B) Kcnmb4 / control, (C) WT Ϫ/Ϫ ␣ KA, and (D) KA Kcnmb4 showing CCD immunohistochemically co-stained with V-ATPase (red) and the 1 subunit of the Na-K- ATPase (green). As shown in all images and quantitatively in panel E, PCs (V-ATPase negative) express strong Na-K-ATPase staining on ␣ the BLM, which increased by 50% in KA. There was no difference in 1 staining between phenotypes on the same diet. E A B

C D F

Ϫ Ϫ Ϫ Ϫ Figure 6. Kcnmb4 / CNT and CCD have blunted distention during KA. Representative sections from (A) WT control, (B) Kcnmb4 / Ϫ Ϫ control, (C) WT KA, and (D) KA Kcnmb4 / illustrating the luminal cross-sectional areas of CCD. Bar plots summarizing the (E) CCD and (F) CNT luminal cross-sectional areas for each treatment group. * ϭ significant difference (P Ͻ 0.05) between diets. † ϭ significant difference between phenotypes on same diet. diet, ICs of CCD and CNT from WT exhibited significant vol- shown in Figure 8, the luminal cross-sectional areas of the PCT ume reduction. However, ICs from KA Kcnmb4Ϫ/Ϫ did not were similarly increased in the KA WT and KA Kcnmb4Ϫ/Ϫ exhibit cell volume reduction. These results indicate that the compared with control. Because BK-␣/␤4 is not normally ex- BK-␤4 of ICs is necessary for high flow, shear-stress-induced pressed in the PCT, this result is not surprising. The TAL, volume decrease. which normally expresses BK-␣/␤4,25 responded similarly as The cross-sectional areas of the proximal tubule (PCT) and the CNT and CCD, with a blunted increase in luminal cross- the TAL were also measured in the four treatment groups. As sectional area in KA Kcnmb4Ϫ/Ϫ compared with KA WT.

638 Journal of the American Society of Nephrology J Am Soc Nephrol 21: 634–645, 2010 www.jasn.org BASIC RESEARCH

secretion with additional epithelial Na channels and Na-K- ATPase in PCs. However, we show that KA Kcnmb4Ϫ/Ϫ exhibit impaired K and Na excretion, yielding volume retention and a reduction in the ratio of K excreted to Na reabsorbed. This was an interesting finding because BK-␣/␤4 resides in ICs, which have minimal Na-K-ATPase, even in KA conditions. Moreover, KA WT and Kcnmb4Ϫ/Ϫ exhibited equal concentrations of plasma aldosterone production and equally increased amounts of Na-K-ATPase in the BLM of PCs. The minimal amount of Na-K-ATPase in ICs suggests that the BK- ␣/␤4 do not have a K secretory role. In addition to the driving force effects of aldosterone, K secretion is enhanced in KA mammals by virtue of increased distal flow rates, which stimulate BK-mediated K secretion. Although the KA Kcnmb4Ϫ/Ϫ accommodated the more than 4-fold increase in flow, the distal tubules remained narrowed and the IC size remained large compared with WT. We con- clude that in KA mice, BK-␣/␤4 play a role in the cell volume reduction of ICs and increase the luminal cross-sectional area in response to the high flow rates through the CNT and CCD. By reducing IC size, the tubule volume is increased, thereby maintaining high chemical gradients for K secretion and reduced chemical gradients for Na reabsorption.

؊/؊ Ϫ/Ϫ Na, K, and Volume Balance in KA Kcnmb4 Figure 7. ICs from Kcnmb4 mice do not undergo flow mediated cell volume decrease. Summary bar plots of measured cell It is not new that BK components are involved in electrolyte 4,26–28 ␣ ␤ volumes of ICs from (A) CCD and (B) CNT determined in each and volume homeostasis. Although the BK- / 4 resides Ϫ Ϫ treatment group. ICs from KA WT, but not KA Kcnmb4 / ,ex- in the apical membrane of ICs, there was a large decrease in the hibited significantly decreased cell volumes in CCD and CNT. * ϭ amount of K secreted and an increase in the amount of Na Ϫ Ϫ significant difference (P Ͻ 0.05) between diets. † ϭ significant absorbed in KA Kcnmb4 / . The K secretory defect in Kc- difference between phenotypes on same diet. nmb4Ϫ/Ϫ was manifested by a reduced urinary K clearance and an increased plasma K concentration. The increased Na and It is conceivable that edema, caused by the fluid retention, water retention was manifested as reduced Na excretion and was causing the constriction of the CNT and CCD of the KA extreme volume expansion. Kcnmb4Ϫ/Ϫ. However, we found no significant difference in In a previous study from this laboratory, we found that KA kidney weights, with values of 149 Ϯ 4, 133 Ϯ 5, 162 Ϯ 3, and Kcnmb1Ϫ/Ϫ (BK-␤1) manifested a K secretory defect, leading Ϫ Ϫ 140 Ϯ 6 mg for control WT, control Kcnmb4 / , KA WT, and to an elevated plasma K concentration and aldosteronism.27 KA Kcnmb4Ϫ/Ϫ, respectively. That the kidney weight did not Another study revealed that BK-␣Ϫ/Ϫ exhibited extreme aldo- Ϫ Ϫ increase in KA Kcnmb4 / indicates the narrowed lumens steronism.6 In the study presented here, the plasma K concen- were not due to renal edema-induced tubular compression. tration increased significantly in the KA Kcnmb4Ϫ/Ϫ; however, The lack of narrowing of the PCT in KA Kcnmb4Ϫ/Ϫ also ar- we found no difference in plasma aldosterone compared with gues against edema-induced tubular compression. KA WT. Aldosterone was probably not elevated in KA Kc- It was determined whether a reduction of BK-␣ expression nmb4Ϫ/Ϫ because of opposing hormonal effects. Because there was the reason for the failure of the CNT and CCD to exhibit a was considerable volume retention in KA Kcnmb4Ϫ/Ϫ, sup- reduction in cell size in the KA Kcnmb4Ϫ/Ϫ. However, as pressed plasma angiotensin II levels would lower aldosterone shown in Supplemental Figure 1, we found no significant dif- production. Moreover, aldosterone was probably suppressed ference in expressions of BK-␣ on ICs (identified by V-AT- by elevated levels of atrial natriuretic factor, which inhibits Pase) of the four treatment groups. adrenal aldosterone production.29–31 The reduced K secretion in the KA Kcnmb4Ϫ/Ϫ can be best explained by the failure to increase the tubular fluid volume. A DISCUSSION reduced tubular volume would result in faster elevation of tubular K concentration as K is secreted in the CNT and CCD. An Normally, KA mammals maintain blood volume and plasma K elevated luminal K concentration reduces the chemical gradi- levels within normal limits by mechanisms including miner- ent for K secretion.32–34 alocorticoid-induced augmentation of the driving force for K The increased Na and fluid retention is not due to primary

J Am Soc Nephrol 21: 634–645, 2010 Role for IC in Na and K Handling 639 BASIC RESEARCH www.jasn.org

C D F

Ϫ Ϫ Figure 8. Kcnmb4 / TAL tubules, but not PCT, have blunted distention during KA. Representative sections from (A) control WT, (B) Ϫ Ϫ Ϫ Ϫ control Kcnmb4 / , (C) KA WT, and (D) KA Kcnmb4 / illustrating the luminal cross-sectional areas of CCD. Bar plots summarizing the (E) PCT and (F) TAL luminal cross-sectional areas for each treatment group. * ϭ significant difference (P Ͻ 0.05) between diets. † ϭ significant difference between phenotypes on same diet. hyperaldosteronism because the plasma aldosterone levels plenishment of the Na and Cl reabsorptive gradient as Na is were the same in the KA WT and KA Kcnmb4Ϫ/Ϫ. It has been extracted in the distal nephron. Therefore, the high shear stress shown that BK-␤4 is predominantly expressed in the brain and of the high velocity through narrowed lumens of KA Kc- Ϫ Ϫ alcohol modulates BK-␣/␤4.35 Therefore, a neural defect of nmb4 / results in increased Na and Cl reabsorption with a Kcnmb4Ϫ/Ϫ, in which the mice are drinking excessively, less negative transepithelial potential and reduced driving should be considered. We observed a moderate but significant force for K secretion, causing the observed volume retention increase in water consumption in the Kcnmb4Ϫ/Ϫ. However, a and rise in plasma K concentration. primary neural defect of excessive drinking by Kcnmb4Ϫ/Ϫ Several other unknown factors released by shear stress may should manifest as hyponatremia and increased urinary vol- be responsible for the reduced K secretion and enhanced Na ume, even in mice on control diets. The normal plasma Na reabsorption in the narrowed tubules of KA Kcnmb4Ϫ/Ϫ. The concentration and decreased urinary output of KA Kc- rapid velocities of fluid would have additional shear stress ef- nmb4Ϫ/Ϫ is consistent with the finding of a primary renal de- fects on ICs that are still protruding into the lumen. Various fect of Na and water retention. compounds may be released from ICs that could inhibit K The normal plasma Na concentration with Na retention for secretion and stimulate Na reabsorption. KA Kcnmb4Ϫ/Ϫ can be explained by an intrinsic renal defect Several studies have shown that shear stress affects the clau- characterized by the failure to increase the tubular cross-sec- dins of tight junctions of endothelial cells.37,38 It is therefore tional area and fluid volume. It has been shown that a reduced possible that paracellular Cl transport is increased to short- cross-sectional area and increased fluid velocity and shear circuit the Na reabsorption and uncouple its exchange for K stress would increase epithelial Na channel-mediated Na reab- secretion. As a molecular switch to control NaCl reabsorption sorption.36 In that study, the increased tubular flow resulted in as opposed to Na-K exchange, with no lysine kinase 4 (WNK4) elevated Na reabsorption without affecting the transepithelial resides in the tight junction of the CCD where it increases potential, indicating that Na transport electrically couples with paracellular Cl transport.39 It would be interesting to deter- paracellular Cl reabsorption rather than K secretion. In addi- mine the effects of shear stress on WNK4 as a regulator of the tion, the faster velocity of flow but with reduced amount of paracellular Cl conductivity of the CNT or CCD. tubular fluid volume in KA Kcnmb4Ϫ/Ϫ results in a faster re- It should be noted that the fluid retention of the KA Kc-

640 Journal of the American Society of Nephrology J Am Soc Nephrol 21: 634–645, 2010 www.jasn.org BASIC RESEARCH

nmb4Ϫ/Ϫ, noted by a weight gain of5gmorethan WT, is dispro- Qϭ␮A (1) portionately greater than indicated by the decrease in hematocrit ␮ (from approximately 45% to 35%). The additional weight is due where is the mean fluid velocity and A is the tubule cross- to extravascular edema because we have observed fluid accumu- sectional area. Under normal and high flow conditions, the lation in the abdominal cavities of the KA Kcnmb4Ϫ/Ϫ. CNT and CCD are mostly impermeable to water. Therefore, one can assume that the flow rate (Q) in these segments is Effects of a High-K Diet on Na-K-ATPase in Distal constant and changes in tubule cross-sectional area will affect fluid velocity (eq 1). On the basis of CNT and CCD cross- Tubular Cells Ϫ/Ϫ Several studies have shown that the amount of Na-K-AT- sectional areas measured in WT and Kcnmb4 in control Pase in cells of the distal nephron is increased in KA condi- and KA conditions (Figure 6), we calculated their average fluid tions.40–42 However, ICs have a paucity of Na-K-AT- velocities (Table 1) on the basis of previously observed flow 49 Pase,20,21 a component necessary for the sustained delivery rates. Paradoxically, in KA WT (i.e., high tubule flow), the of K through the cell and to the apical BK-␣/␤4. Studies velocity decreases substantially because of the large increase in have not determined whether the Na-K-ATPase of ICs is tubule cross-sectional area. This enhances distal tubule fluid also increased under KA conditions. These results support volume and reduces the luminal K concentration, allowing a the finding that the Na-K-ATPase of ICs is Ͻ10% of the more favorable chemical gradient for K secretion (Figure 1B), thereby explaining the reduced K excretion in the KA Kc- quantity in the PCs or CNT. Although we found an increase Ϫ/Ϫ in staining intensity of Na-K-ATPase in the BLM of the PCs nmb4 and the BK-mediated flow-induced K secretion in 9,50 and CNT of KA mice, we did not find an increase in the Na-K- the isolated perfused CCD. ATPase of the ICs. The Na-K-2Cl transporter (NKCC), de- Changes in tubular cross-sectional area also can be ex- scribed in MDCK-C11 cells and which has many IC properties, plained by the hydrostatic pressure gradients along the distal is a potential source of K entering the cell via the basolateral nephron that are necessary to drive fluid flow. The Hagen– membrane.43 The NKCC could serve to replenish the K of the Poiseuille equation describes the volumetric flow rate (Q)in ␮ ICs when flow ceases and the cells return to their normal size. terms of tubule radius (a), fluid viscosity ( ), and hydrostatic However, for sustained K transport via NKCC, a primary ac- pressure gradient (dp/dz): tive Na pump other than the Na-K-ATPase found in the PCs Ϫ␲a4 dp Q ϭ (2) would be necessary to maintain a low intracellular Na concen- 8␮ dz tration. These results suggest that, unlike the PCs, the ICs are not directly engaged in net K transport of KA mice. Therefore, Because ␮ is constant, Q is a linear function of pressure drop the source of the enhanced K secretion in the distal nephron of but a quadratic function of tubule radius. Therefore, small KA mammals is likely in the CNT and PCs, which exhibit a changes in tubule radius can profoundly affect the volumetric 50% increase in Na-K-ATPase. flow rate. For the given flow rates (Table 1), we plotted the required hydrostatic pressure gradient (dp/dz), for a given tu- Regulation of IC Volume by BK-␣/␤4 bule radius (Figure 9). Radii were chosen based on the range of Our data showed that the ICs from KA WT, but not KA Kc- cross-sectional areas that were measured (Figure 6) assuming a Ϫ Ϫ nmb4 / , exhibited a reduction of cell size. This result is consis- circular cross-section (A ϭ ␲a2). Assuming that the CNT did tent with high distal flow activating BK-␣/␤4, with the BK-␤4asa not dilate under high flow conditions (a ϭ 5 ␮m), then a pres- necessary component, initiating a volume regulatory decrease. sure gradient of 25 mmHg/cm would be needed to maintain a ICs are bestowed with a relatively high cell potential of approxi- 7-nl/min flow. Because the hydrostatic pressure upstream in mately Ϫ30 mV44,45 because of a large dominating basolateral Cl Bowman’s space is approximately 15 mmHg, this would not be conductance.46 Thus, there is a large electrochemical gradient for possible. However, increasing tubule radius to only 7 ␮m K to exit the cell when K channels are activated. In this scenario, on would require a more feasible pressure gradient of only 6.6 shear-stress activation of BK-␣/␤4, the intracellular K would exit mmHg/cm. As tubule radius increases, the need for large pres- the cell, Cl would follow the electrical gradient, and water would exit down an osmotic gradient via aquaporin 6 channels, which Table 1. Average fluid velocitya 47 were previously identified in ICs. Indeed, we have recently Fluid Velocity (mm/min) CNTb CCDc shown that shear stress activates BK-␣/␤4 of MDCK-C11 cells, an WT, normal flow 23.44 21.28 IC clone of distal tubule cells, causing a reduction in cell K con- Ϫ Ϫ Kcnmb4 / , normal flow 22.39 20.80 tent. This effect was eliminated by transfecting MDCK-C11 with WT, high flow 15.77 25.05 ␤ 48 Ϫ Ϫ BK- 4 small interfering RNA. Kcnmb4 / , high flow 25.18 38.82 aFluid velocities in the CNT and CCD for given volumetric flow rates were Reduction of IC Size Affects Tubular Flow calculated based on eq 1 and the CNT and CCD cross-sectional areas given The volumetric flow rate (Q) of an incompressible, fully devel- in Figure 6. Animals on a control diet exhibited normal tubular flow. Animals on a high-K diet exhibited high flow oped fluid through a tube of any cross-sectional geometry can bCNT: normal flow ϭ 3 nl/min, high flow ϭ 7 nl/min. be expressed as cCCD: normal flow ϭ 15 nl/min, high flow ϭ 35 nl/min.

J Am Soc Nephrol 21: 634–645, 2010 Role for IC in Na and K Handling 641 BASIC RESEARCH www.jasn.org

CONCISE METHODS

Animal Studies All animals were maintained in accordance with the Institutional An- imal Care and Use Committee of the University of Nebraska Medical Center. At approximately 8 weeks of age, WT (C57BL/6, Charles River, Wilmington, MA) and Kcnmb4Ϫ/Ϫ mice (generously provided by R. Brenner) were given normal mouse chow (0.6% Kϩ, 0.32% Naϩ, control) or chow with a high K content (5.0% Kϩ, 0.32% Naϩ, from Harlan Teklad, Madison, WI), for 10 days. The animals had full access to water at all times. Urine samples were collected several times a day using metabolic cages (Nalgene), as described previously.27 The Kcnmb4Ϫ/Ϫ mice were developed and verified by tissue-typ- ing by Brenner et al.52 The genotype was further verified in our labo- Ϫ Ϫ Figure 9. High pressure gradients are required in the absence of ratory by the lack of BK-␤4 protein in renal cortex of Kcnmb4 / .25 tubule dilation. Pressure gradient (dp/dz) was plotted as a func- After treatment, mice were anesthetized and blood samples were tion of tubule radius for a given volumetric flow rate in the CNT harvested as described previously27 for measuring hematocrit and and CCT according to the Hagen–Poiseuille equation (eq 2). plasma aldosterone, creatinine, Na, and K concentrations. A flame photometer (Jenway model PFP7/C, Barloworld Scientific Ltd., Essex, sure gradients at high flow rates decreases substantially (a func- 4 United Kingdom) was used to measure plasma and urine Na and K tion of a ), making this less of an issue in the CCD. concentrations. Plasma aldosterone (Cayman Chemical, Ann Arbor, Relative to PCs, ICs protrude into the lumen, acting as “speed MI) and creatinine (BioAssay Systems, Hayward, CA) concentrations bumps” along the distal nephron, disturbing the tubule fluid flow were measured by colorimetric assay following the manufacturer’s field, and increasing tubular resistance to flow. Flow-induced BK protocol. Kidneys were immediately fixed in Bouins solution (Lab- activation in ICs leads to a substantial reduction in IC volume, Chem, Pittsburg, PA) or Histochoice MB (Electron Microscopy Sci- thereby reducing their luminal profile, decreasing tubular resis- ences, Hatfield, PA) and embedded in paraffin for sectioning as de- tance, and increasing tubular volumetric fluid flow. scribed previously.25,53 BK-␤4 was also found in the distal convoluted tubule and the TAL.25 BK are thought to have a role in cell volume regulation in TAL cells.51 The cross-sectional areas of the distal Immunohistochemical Staining convoluted tubule were not measured in this study. However, For colorimetric immunohistochemical staining, kidney sections 4 consistent with a role for BK-␣/␤4 in reducing cell size with ␮m thick were processed at room temperature according to stan- increased flow, we detected significantly lower TAL cross-sec- dard immunohistochemical methods as described previously.28 Ϫ Ϫ tional areas in KA Kcnmb4 / . We found no difference in Briefly, paraffin slides were washed in xylene and then rehydrated cross-sectional areas of the PCT, which does not have BK␤4, by a series of ethanol washes. When necessary, antigen retrieval Ϫ Ϫ when KA WT was compared with KA Kcnmb4 / . was performed by incubating slides for 5 minutes at room temperature in PBS with 1% SDS, followed by washing in PBS. Endoge- Summary and Relevance nous peroxidase was quenched with 10% hydrogen peroxide for 20 The results of this study provide a different view on the role of minutes followed by permeabilization with 0.5% Triton X-100 for BK in ICs with respect to flow-mediated K secretion elicited by 20 minutes and blocking (PBS with 1% BSA and 1% powdered a high-K diet. The finding that PCs, but not ICs, of KA mice milk) for 30 minutes. Sections were incubated with the primary exhibit an increase in basolateral Na-K-ATPase suggests that mouse antibodies diluted in blocking buffer overnight at 4°C. After- ICs may not transport K transcellularly. Instead, the high flow wards, sections were washed three times in PBS with 0.5% Tween-20 and Ј conditions of K adaptation cause a decrease in IC size and an incubated with either species-specific F(ab )2-horseradish-peroxidase- Ј increase in luminal diameter, resulting in larger volumes of conjugated or F(ab )2-alkaline-phosphatase-conjugated or both second- distal fluid. A reduced K concentration in the CNT and CCD ary antibodies (Santa Cruz Biotechnology) for 1 hour at room tempera- yields a more favorable chemical gradient for secreting K. That ture. Sections were rinsed three times in PBS with 0.5% Tween-20 and the PCs exhibited an increase in Na-K-ATPase in the KA mice then developed with either diaminobenzidin substrate or AP-Fast Red suggests that the PCs are secreting K transcellularly, probably (both from Invitrogen) following the manufacturer’s protocol. CNT through the BK-␣/␤1 (in CNT) or ROMK. The BK-␣/␤4of were identified by basolateral staining with a Na-Ca exchanger mouse ICs may take part in sustained K secretion in KA conditions monoclonal antibody (Swant, Switzerland, 1:200 dilution), and CCD through an active transport mechanism that does not involve were identified by basolateral staining with a goat polyclonal aquaporin 3 the Na-K-ATPase. However, the data of this study show that an antibody (Santa Cruz Biotechnology, 1:50 dilution). Similarly, PCT and important component of flow-mediated K secretion is the re- TAL were identified by staining with goat polyclonal Na/glucose cotrans- duction of IC size via activation of BK-␣/␤4. porter antibody (Santa Cruz Biotechnology, 1:50 dilution) and rabbit

642 Journal of the American Society of Nephrology J Am Soc Nephrol 21: 634–645, 2010 www.jasn.org BASIC RESEARCH

polyclonal Tamm–Horsfall glycoprotein (Santa Cruz Biotechnology, HK) for WT and Kcnmb4Ϫ/Ϫ mice were determined by ANOVA plus 1:50 dilution), respectively. Student-Newman-Keuls or tukey test (P Ͻ 0.05 considered signifi- For fluorescence immunohistochemical staining, kidney sections cant). The coefficient of error for the Cavalieri method was deter- were processed as described above without peroxidase quenching. mined using the subsampling method.24 We performed data manage- Sections were incubated with rabbit polyclonal anti-LEAVE (gener- ment and statistical analyses using Excel (Microsoft Corporation, ␣ ous gift from Dr. Thomas Pressley), a specific antibody for the 1 Redmond, WA) and SigmaPlot (version 11, Systat Software, GmbH, subunit of the Na-K-ATPase54 (1:100 dilution), or mouse monoclo- Germany). nal anti-BK␣ (NeuroMab, University of California, Davis, CA; diluted 1:50) and goat polyclonal anti-VATPase (diluted 1:100, Santa Cruz Biotechnology, Santa Cruz, CA) in blocking buffer overnight at ACKNOWLEDGMENTS 4°C. Rabbit or mouse IgG was used as a negative control. After washing three times in PBS with 0.5% Tween-20, sections were incubated This work was supported by National Institutes of Diabetes and Di- for 1 hour in the dark at room temperature with either donkey anti- gestive and Kidney Diseases grants RO1 DK49461 and RO1 DK73070 rabbit IgG conjugated Alexa Fluor 488 or donkey anti-mouse IgG (to S.C. Sansom) and a fellowship (0610059Z) from the American conjugated Alexa Fluor 488 and donkey anti-goat IgG conjugated Heart Association-Heartland Affiliate (P.R. Grimm). The monoclo- Alexa Fluor 594 (both diluted 1:200 in blocking buffer, Invitrogen). nal antibody against BK-␣ was developed by and/or obtained from ␮ After nuclear staining with 0.25 g/ml Hoechst 33258 for 10 minutes the University of California–Davis/National Institutes of Health in the dark at room temperature, sections were rinsed three times in (NIH) NeuroMab Facility, supported by NIH Grant U24-NS-050606 PBS with 0.5% Tween-20, dried, mounted with Prolong Gold (In- and maintained by the Department of Neurobiology, Physiology, and vitrogen), sealed with nail polish, and viewed on a Leica HC fluores- Behavior, College of Biologic Sciences, at the University of California ϫ cence microscope with a 40 /0.75 NA HCX PL Fluotar objective with in Davis, CA. We thank Tom Barger of the Electron Microscopy Re- images captured with an QImaging Retiga EXi CCD camera (Surrey, search Facility at the University of Nebraska Medical Center for his British Columbia, Canada) and analyzed with ImageJ software (ver- assistance and technical advice. A portion of this work was presented sion 1.42, National Institutes of Health, Bethesda, MD). Quantifica- at the annual Experimental Biology Meeting, April 5 to 9, 2008, San tion of basolateral Na-K-ATPase signal intensity in PCs (V-ATPase Diego, CA, in abstract form. negative) and ICs (V-ATPase positive) was determined following online instructions in single-channel, gray scale images after background correction. DISCLOSURES None. Measurements of Tubule Cross-Sectional Area Tubule cross-sectional area was measured by counting total pixels ϭ within the tubule and converting pixel size to area (1 pixel 0.14 REFERENCES ␮m2). All images were taken at the same magnification and resolution ϫ on a Leica HC fluorescence microscope with a 40 /0.75 NA HCX PL 1. Battilana CA, Dobyan DC, Lacy FB, Bhattacharya J, Johnston PA, Fluotar objective with images captured using a high-resolution digital Jamison RL: Effect of chronic potassium loading on potassium secre- CCD color camera (QImaging MicroPublisher 3.3 RTV, Surrey, Brit- tion by the pars recta or descending limb of the juxtamedullary ish Columbia, Canada) and analyzed with Photoshop (version 6.2). nephron in the rat. J Clin Invest 62: 1093–1103, 1978 Ten PCT, TAL, CNT, and CCD tubules were randomly selected per 2. Stokes JB: Consequences of potassium recycling in the renal medulla. Effects of ion transport by the medullary thick ascending limb of section. Four nonserial sections were measured per mouse, and four Henle’s loop. J Clin Invest 70: 219–229, 1982 mice per group, giving 160 tubule cross-sectional area measurements 3. Grimm PR, Sansom SC: BK channels in the kidney. Curr Opin Nephrol per tubular segment per group. Hypertens 16: 430–436, 2007 4. Bailey MA, Cantone A, Yan Q, MacGregor GG, Leng Q, Amorim JB, Wang T, Hebert SC, Giebisch G, Malnic G: Maxi-K channels contribute IC Volume Measurements to urinary potassium excretion in the ROMK-deficient mouse model of IC volume was estimated using the well established Cavalieri method Type II Bartter’s syndrome and in adaptation to a high-K diet. Kidney as described previously.24 Four serial sections 3 ␮m thick were stained Int 70: 51–59, 2006 as described above. Ten CNT and ten CCD were randomly selected 5. Sansom SC, Welling PA: Two channels for one job. Kidney Int 72: and ICs were identified in each tubule by the lack of basolateral stain- 529–530, 2007 6. Rieg T, Vallon V, Sausbier M, Sausbier U, Kaissling B, Ruth P, ing of the Na-Ca exchanger 55,56 or aquaporin 3.57 IC cross-sectional Osswald H: The role of the BK channel in potassium homeostasis area was measure as described above for each cell. The IC volume was and flow-induced renal potassium excretion. Kidney Int 72: 566– estimated by summing the IC cross-sectional areas and multiplying by 573, 2007 the slice thickness. At least 40 IC volumes were measured per group. 7. Lu M, Wang T, Yan Q, Yang X, Dong K, Knepper MA, Wang W, Giebisch G, Shull GE, Hebert SC: Absence of small conductance Kϩ channel (SK) activity in apical membranes of thick ascending limb and Statistical Methods cortical collecting duct in ROMK (Bartter’s) knockout mice. J Biol Ϯ Data shown in figures represent mean SEM. Unless otherwise de- Chem 277: 37881–37887, 2002 noted, significant differences between treatment groups (control, 8. Kunau RT Jr, Webb HL, Borman SC: Characteristics of the relationship

J Am Soc Nephrol 21: 634–645, 2010 Role for IC in Na and K Handling 643 BASIC RESEARCH www.jasn.org

between the flow rate of tubular fluid and potassium transport in the in Naϩ reabsorption by cortical collecting ducts of Naϩ-deprived distal tubule of the rat. J Clin Invest 54: 1488–1495, 1974 mice. Am J Physiol Renal Physiol 297: F420–F428, 2009 9. Woda CB, Bragin A, Kleyman TR, Satlin LM: Flow-dependent Kϩ 29. Barrett PQ, Isales CM, Bollag WB, McCarthy RT: Ca2ϩ channels and secretion in the cortical collecting duct is mediated by a maxi-K aldosterone secretion: Modulation by Kϩ and atrial natriuretic pep- channel. Am J Physiol Renal Physiol 280: F786–F793, 2001 tide. Am J Physiol 261: F706–F719, 1991 10. Palmer LG, Frindt G: High-conductance K: Channels in intercalated 30. Szalay KS, Beck M, Toth M, de CR: Interactions between ouabain, cells of the rat distal nephron. Am J Physiol Renal Physiol 292: F966– atrial natriuretic peptide, angiotensin-II and potassium: Effects on rat F973, 2007 zona glomerulosa aldosterone production. Life Sci 62: 1845–1852, 11. Pacha J, Frindt G, Sackin H, Palmer LG: Apical maxi K channels in 1998 intercalated cells of CCT. Am J Physiol 261: F696–F705, 1991 31. Isales CM, Bollag WB, Kiernan LC, Barrett PQ: Effect of ANP on 12. Pech V, Kim YH, Weinstein AM, Everett LA, Pham TD, Wall SM: sustained aldosterone secretion stimulated by angiotensin II. Am J Angiotensin II increases chloride absorption in the cortical collecting Physiol 256: C89–C95, 1989 duct in mice through a pendrin-dependent mechanism. Am J Physiol 32. Good DW, Wright FS: Luminal influences on potassium secretion: Renal Physiol 292: F914–F920, 2007 Sodium concentration and fluid flow rate. Am J Physiol 236: F192– 13. Li D, Wang Z, Sun P, Jin Y, Lin DH, Hebert SC, Giebisch G, Wang WH: F205, 1979 Inhibition of MAPK stimulates the Ca2ϩ -dependent big-conductance 33. Wright FS: Flow-dependent transport processes: Filtration, absorp- K channels in cortical collecting duct. Proc Natl Acad Sci U S A 103: tion, secretion. Am J Physiol 243: F1–11, 1982 19569–19574, 2006 34. Engbretson BG, Stoner LC: Flow-dependent potassium secretion by 14. Le HM, Kaissling B, Dubach UC: Distal tubular segments of the rabbit rabbit cortical collecting tubule in vitro. Am J Physiol 253: F896–F903, kidney after adaptation to altered Na- and K-intake. II. Changes in 1987 Na-K-ATPase activity. Cell Tissue Res 224: 493–504, 1982 35. Martin GE, Hendrickson LM, Penta KL, Friesen RM, Pietrzykowski AZ, 15. Garg LC, Knepper MA, Burg MB: Mineralocorticoid effects on Na-K- Tapper AR, Treistman SN: Identification of a BK channel auxiliary ATPase in individual nephron segments. Am J Physiol 240: F536– protein controlling molecular and behavioral tolerance to alcohol. F544, 1981 Proc Natl Acad SciUSA105: 17543–17548, 2008 16. Tsuchiya K, Giebisch G, Welling PA: Aldosterone-dependent regula- 36. Satlin LM, Sheng S, Woda CB, Kleyman TR: Epithelial Na(ϩ) channels tion of Na-K-ATPase subunit mRNA in the rat CCD: Competitive PCR are regulated by flow. Am J Physiol Renal Physiol 280: F1010–F1018, analysis. Am J Physiol 271: F7–F15, 1996 2001 17. Hayhurst RA, O’Neil RG: Time-dependent actions of aldosterone and 37. DeMaio L, Chang YS, Gardner TW, Tarbell JM, Antonetti DA: Shear amiloride on Naϩ-Kϩ-ATPase of cortical collecting duct. Am J Physiol stress regulates occludin content and phosphorylation. Am J Physiol 254: F689–F696, 1988 Heart Circ Physiol 281: H105–H113, 2001 18. Ridderstrale Y, Kashgarian M, Koeppen B, Giebisch G, Stetson D, 38. Colgan OC, Ferguson G, Collins NT, Murphy RP, Meade G, Cahill Ardito T, Stanton B: Morphological heterogeneity of the rabbit col- PA, Cummins PM: Regulation of bovine brain microvascular en- lecting duct. Kidney Int 34: 665–670, 1988 dothelial tight junction assembly and barrier function by laminar 19. Holthofer H, Schulte BA, Pasternack G, Siegel GJ, Spicer SS: Three shear stress. Am J Physiol Heart Circ Physiol 292: H3190–H3197, distinct cell populations in rat kidney collecting duct. Am J Physiol 2007 253: C323–C328, 1987 39. Kahle KT, MacGregor GG, Wilson FH, Van Hoek AN, Brown D, Ardito 20. Sabolic I, Herak-Kramberger CM, Breton S, Brown D: Na/K-ATPase in T, Kashgarian M, Giebisch G, Hebert SC, Boulpaep EL, Lifton RP: intercalated cells along the rat nephron revealed by antigen retrieval. Paracellular ClϪ permeability is regulated by WNK4 kinase: Insight J Am Soc Nephrol 10: 913–922, 1999 into normal physiology and hypertension. Proc Natl Acad SciUSA 21. Sampaio MS, Bezerra IP, Pecanha FL, Fonseca PH, Capella MA, Lopes 101: 14877–14882, 2004 AG: Lack of Na(ϩ),K (ϩ)-ATPase expression in intercalated cells may 40. Fujii Y, Mujais SK, Katz AI: Renal potassium adaptation: Role of the be compensated by Na(ϩ)-ATPase: A study on M. Cell Mol Life Sci 65: Naϩ-Kϩ pump in rat cortical collecting tubules. Am J Physiol 256: 3093–3099, 2008 F279–F284, 1989 22. Pluznick JL, Sansom SC: BK channels in the kidney: Role in K(ϩ) 41. Garg LC, Narang N: Effects of potassium bicarbonate on distal secretion and localization of molecular components. Am J Physiol nephron Na-K-ATPase in adrenalectomized rabbits. Pfugers Arch 409: Renal Physiol 291: F517–F529, 2006 126–131, 1987 23. Feraille E, Mordasini D, Gonin S, Deschenes G, Vinciguerra M, Doucet 42. Palmer LG, Antonian L, Frindt G: Regulation of apical K and Na A, Vandewalle A, Summa V, Verrey F, Martin PY: Mechanism of control channels and Na/K pumps in rat cortical collecting tubule by dietary K. of Na-K-ATPase in principal cells of the mammalian collecting duct. J Gen Physiol 104: 693–710, 1994 Ann NY Acad Sci 986: 570–578, 2003 43. Akimova OA, Taurin S, Dulin NO, Orlov SN: Purinergic inhibition of 24. Howard MA, Roberts N, Garcia-Finana M, Cowell PE: Volume estima- Na(ϩ),K (ϩ),Cl (Ϫ) cotransport in C11-MDCK cells: Role of stress- tion of prefrontal cortical subfields using MRI and stereology. Brain activated protein kinases. Purinergic Signal 4: 183–191, 2008 Res Brain Res Protoc 10: 125–138, 2003 44. Muto S, Giebisch G, Sansom S: Effects of adrenalectomy on CCD: 25. Grimm PR, Foutz RM, Brenner R, Sansom SC: Identification and local- Evidence for differential response of two cell types. Am J Physiol 253: ization of BK-beta subunits in the distal nephron of the mouse kidney. F742–F752, 1987 Am J Physiol Renal Physiol 293: F350–F359, 2007 45. Koeppen BM: Electrophysiological identification of principal and in- 26. Sausbier M, Arntz C, Bucurenciu I, Zhao H, Zhou XB, Sausbier U, Feil tercalated cells in the rabbit outer medullary collecting duct. Pflugers S, Kamm S, Essin K, Sailer CA, Abdullah U, Krippeit-Drews P, Feil R, Archiv European J Physiol 409: 138–141, 1987 Hofmann F, Knaus HG, Kenyon C, Shipston MJ, Storm JF, Neuhuber 46. Muto S, Yasoshima K, Yoshitomi K, Imai M, Asano Y: Electrophysio- W, Korth M, Schubert R, Gollasch M, Ruth P: Elevated blood pressure logical identification of alpha- and beta-intercalated cells and their linked to primary hyperaldosteronism and impaired vasodilation in BK distribution along the rabbit distal nephron segments. J Clin Invest 86: channel-deficient mice. Circulation 112: 60–68, 2005 1829–1839, 1990 27. Grimm PR, Irsik DL, Settles DC, Holtzclaw JD, Sansom SC: Hyperten- 47. Ohshiro K, Yaoita E, Yoshida Y, Fujinaka H, Matsuki A, Kamiie J, sion of Kcnmb1Ϫ/Ϫ is linked to deficient K secretion and aldosteron- Kovalenko P, Yamamoto T: Expression and immunolocalization of ism. Proc Natl Acad SciUSA106: 11800–11805, 2009 AQP6 in intercalated cells of the rat kidney collecting duct. Arch Histol 28. Grimm PR, Irsik DL, Liu L, Holtzclaw JD, Sansom SC: Role of BKbeta1 Cytol 64: 329–338, 2001

644 Journal of the American Society of Nephrology J Am Soc Nephrol 21: 634–645, 2010 www.jasn.org BASIC RESEARCH

48. Holtzclaw JD, Rigle JD, Irsik DL, Grimm PR, and Sansom, SC: The BK 54. Pressley TA: Phylogenetic conservation of isoform-specific regions channel in intercalated cells mediates shear stress induced ATP de- within alpha-subunit of Na(ϩ)-K(ϩ)-ATPase. Am J Physiol 262: C743– pendent K efflux [Abstract]. FASEB J 23: LB146, 2009 C751, 1992 49. Wu L, Gao X, Brown RC, Heller S, O’Neil RG: Dual role of the TRPV4 55. Loffing J, Loffing-Cueni D, Valderrabano V, Klausli L, Hebert SC, channel as a sensor of flow and osmolality in renal epithelial cells. Rossier BC, Hoenderop JG, Bindels RJ, Kaissling B: Distribution of Am J Physiol Renal Physiol 293: F1699–F1713, 2007 transcellular calcium and sodium transport pathways along mouse 50. Liu W, Morimoto T, Woda C, Kleyman TR, Satlin LM: Ca2ϩ depen- distal nephron. Am J Physiol Renal Physiol 281: F1021–F1027, 2001 dence of flow-stimulated K secretion in the mammalian cortical col- 56. Loffing J, Kaissling B: Sodium and calcium transport pathways along lecting duct. Am J Physiol Renal Physiol 293: F227–F235, 2007 the mammalian distal nephron: from rabbit to human. Am J Physiol 51. Montrose-Rafizadeh C, Guggino WB: Role of intracellular calcium in Renal Physiol 284: F628–F643, 2003 volume regulation by rabbit medullary thick ascending limb cells. 57. Zharkikh L, Zhu X, Stricklett PK, Kohan DE, Chipman G, Breton S, Am J Physiol 260: F402–F409, 1991 Brown D, Nelson RD: Renal principal cell-specific expression of green 52. Brenner R, Chen QH, Vilaythong A, Toney GM, Noebels JL, Aldrich fluorescent protein in transgenic mice. Am J Physiol Renal Physiol 283: RW: BK channel beta4 subunit reduces dentate gyrus excitability and F1351–F1364, 2002 protects against temporal lobe seizures. Nature Neurosci 8: 1752– 1759, 2005 53. Wei P, Lane PH, Lane JT, Padanilam BJ, Sansom SC: Glomerular structural and functional changes in a high-fat diet mouse model Supplemental information for this article is available online at http://www. of early-stage type 2 diabetes. Diabetologia 47: 1541–1549, 2004 jasn.org/.

J Am Soc Nephrol 21: 634–645, 2010 Role for IC in Na and K Handling 645